progan.patch

diff --git a/models/trainer/standard_configurations/pgan_config.py b/models/trainer/standard_configurations/pgan_config.py
index 59ad0ea..77b38e9 100644
--- a/models/trainer/standard_configurations/pgan_config.py
+++ b/models/trainer/standard_configurations/pgan_config.py
@@ -45,7 +45,7 @@ _C.alphaSizeJumps = [0, 32, 32, 32, 32, 32, 32, 32, 32, 32]
 _C.depthScales = [512, 512, 512, 512, 256, 128, 64, 32, 16]
 
 # Mini batch size
-_C.miniBatchSize = 16
+_C.miniBatchSize = 8
 
 # Dimension of the latent vector
 _C.dimLatentVector = 512
diff --git a/models/utils/utils.py b/models/utils/utils.py
index 2a215a7..cef5d9b 100644
--- a/models/utils/utils.py
+++ b/models/utils/utils.py
@@ -341,3 +341,26 @@ def saveScore(outPath, outValue, *args):
         json.dump(fullDict, file, indent=2)
 
     os.remove(flagPath)
+
+
+def load_progan(name='jelito3d_batchsize8', checkPointDir='output_networks/jelito3d_batchsize8', freeze_pgan_disc=True):
+    """Load pretrained ProGAN from checkpoint."""
+    checkpointData = getLastCheckPoint(checkPointDir, name, scale=None, iter=None)
+    modelConfig, pathModel, _ = checkpointData
+    _, scale, _ = parse_state_name(pathModel)
+
+    module = 'PGAN'
+    packageStr, modelTypeStr = getNameAndPackage(module)
+    modelType = loadmodule(packageStr, modelTypeStr)
+
+    with open(modelConfig, 'rb') as file:
+        config = json.load(file)
+
+    model = modelType(useGPU=True, storeAVG=True, **config)
+    model.load(pathModel)
+
+    if freeze_pgan_disc:
+        for param in model.netD.parameters():
+            param.requires_grad = False
+
+    return model
diff --git a/visualization/visualizer.py b/visualization/visualizer.py
index f3aab21..3874796 100644
--- a/visualization/visualizer.py
+++ b/visualization/visualizer.py
@@ -6,7 +6,7 @@ import torchvision.utils as vutils
 import numpy as np
 import random
 
-vis = visdom.Visdom()
+# vis = visdom.Visdom()
 
 
 def resizeTensor(data, out_size_image):